Ensure accepted input types are general where possible (#480)

* Strings with FunctionTransform

* String with SelectTransform

* Strings with ConstantKernel

* Add Vector{String} test util

* Constant kernelswith strings

* Test kernelproduct with strings

* Test kernelsum with strings

* Transformed kernel with string input

* KernelTensorProduct with String

* Bump patch

* Formatting

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

* Formatting

* Formatting suggestions

Co-authored-by: David Widmann <[email protected]>

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: David Widmann <[email protected]>

Author: 3 people

Project.toml

@@ -1,6 +1,6 @@
 name = "KernelFunctions"
 uuid = "ec8451be-7e33-11e9-00cf-bbf324bd1392"
-version = "0.10.46"
+version = "0.10.47"
 
 [deps]
 ChainRulesCore = "d360d2e6-b24c-11e9-a2a3-2a2ae2dbcce4"

src/TestUtils.jl

@@ -142,6 +142,28 @@ function test_interface(
     )
 end
 
+function test_interface(rng::AbstractRNG, k::Kernel, ::Type{Vector{String}}; kwargs...)
+    return test_interface(
+        k,
+        [randstring(rng) for _ in 1:3],
+        [randstring(rng) for _ in 1:3],
+        [randstring(rng) for _ in 1:4];
+        kwargs...,
+    )
+end
+
+function test_interface(
+    rng::AbstractRNG, k::Kernel, ::Type{ColVecs{String}}; dim_in=2, kwargs...
+)
+    return test_interface(
+        k,
+        ColVecs([randstring(rng) for _ in 1:dim_in, _ in 1:3]),
+        ColVecs([randstring(rng) for _ in 1:dim_in, _ in 1:3]),
+        ColVecs([randstring(rng) for _ in 1:dim_in, _ in 1:4]);
+        kwargs...,
+    )
+end
+
 function test_interface(k::Kernel, T::Type{<:AbstractVector}; kwargs...)
     return test_interface(Random.GLOBAL_RNG, k, T; kwargs...)
 end

src/distances/delta.jl

@@ -1,10 +1,9 @@
 # Delta is not following the PreMetric rules since d(x, x) == 1
 struct Delta <: Distances.UnionPreMetric end
 
-(dist::Delta)(a::Number, b::Number) = a == b
-Base.@propagate_inbounds function (dist::Delta)(
-    a::AbstractArray{<:Number}, b::AbstractArray{<:Number}
-)
+(dist::Delta)(a, b) = a == b
+
+Base.@propagate_inbounds function (dist::Delta)(a::AbstractArray, b::AbstractArray)
     @boundscheck if length(a) != length(b)
         throw(
             DimensionMismatch(

src/transform/selecttransform.jl

@@ -28,7 +28,7 @@ _maybe_unwrap(x::AbstractArray{<:Any,0}) = x[]
 _map(t::SelectTransform, x::ColVecs) = _wrap(view(x.X, t.select, :), ColVecs)
 _map(t::SelectTransform, x::RowVecs) = _wrap(view(x.X, :, t.select), RowVecs)
 
-_wrap(x::AbstractVector{<:Real}, ::Any) = x
-_wrap(X::AbstractMatrix{<:Real}, ::Type{T}) where {T} = T(X)
+_wrap(x::AbstractVector, ::Any) = x
+_wrap(X::AbstractMatrix, ::Type{T}) where {T} = T(X)
 
 Base.show(io::IO, t::SelectTransform) = print(io, "Select Transform (dims: ", t.select, ")")

test/basekernels/constant.jl

@@ -8,6 +8,7 @@
 
         # Standardised tests.
         TestUtils.test_interface(k, Float64)
+        TestUtils.test_interface(k, Vector{String})
         test_ADs(ZeroKernel)
         test_interface_ad_perf(_ -> k, nothing, StableRNG(123456))
     end
@@ -22,6 +23,7 @@
 
         # Standardised tests.
         TestUtils.test_interface(k, Float64)
+        TestUtils.test_interface(k, Vector{String})
         test_ADs(WhiteKernel)
         test_interface_ad_perf(_ -> k, nothing, StableRNG(123456))
     end
@@ -38,6 +40,7 @@
 
         # Standardised tests.
         TestUtils.test_interface(k, Float64)
+        TestUtils.test_interface(k, Vector{String})
         test_ADs(c -> ConstantKernel(; c=only(c)), [c])
         test_interface_ad_perf(c -> ConstantKernel(; c=c), c, StableRNG(123456))
     end

test/kernels/kernelproduct.jl

@@ -11,6 +11,7 @@
 
     # Standardised tests.
     TestUtils.test_interface(k, Float64)
+    TestUtils.test_interface(ConstantKernel(; c=1.0) * WhiteKernel(), Vector{String})
     test_ADs(
         x -> KernelProduct(SqExponentialKernel(), LinearKernel(; c=exp(x[1]))), rand(1)
     )

test/kernels/kernelsum.jl

@@ -12,6 +12,7 @@
 
     # Standardised tests.
     TestUtils.test_interface(k, Float64)
+    TestUtils.test_interface(ConstantKernel(; c=1.5) * WhiteKernel(), Vector{String})
     test_ADs(x -> KernelSum(SqExponentialKernel(), LinearKernel(; c=exp(x[1]))), rand(1))
     test_interface_ad_perf(2.4, StableRNG(123456)) do c
         KernelSum(SqExponentialKernel(), LinearKernel(; c=c))

test/kernels/kerneltensorproduct.jl

@@ -32,6 +32,9 @@
     # Standardised tests.
     TestUtils.test_interface(kernel1, ColVecs{Float64})
     TestUtils.test_interface(kernel1, RowVecs{Float64})
+    TestUtils.test_interface(
+        KernelTensorProduct(WhiteKernel(), ConstantKernel(; c=1.1)), ColVecs{String}
+    )
     test_ADs(
         x -> KernelTensorProduct(SqExponentialKernel(), LinearKernel(; c=exp(x[1]))),
         rand(1);

test/kernels/transformedkernel.jl

@@ -25,6 +25,10 @@
     @test repr(kt) == repr(k) * "\n\t- " * repr(ScaleTransform(s))
 
     TestUtils.test_interface(k, Float64)
+    TestUtils.test_interface(
+        TransformedKernel(ConstantKernel(; c=1.5), FunctionTransform(x -> x * "hi")),
+        Vector{String},
+    )
     test_ADs(x -> SqExponentialKernel() ∘ ScaleTransform(x[1]), rand(1))
     test_interface_ad_perf(0.35, StableRNG(123456)) do λ
         SqExponentialKernel() ∘ ScaleTransform(λ)

test/transform/functiontransform.jl

@@ -26,6 +26,15 @@
         end
     end
 
+    @testset "String input" begin
+        f = x -> x * "hello"
+        t = FunctionTransform(f)
+        x = [randstring(rng) for _ in 1:3]
+        y = map(t, x)
+        @test all([t(x[n]) == y[n] for n in eachindex(x)])
+        @test all([f(x[n]) == y[n] for n in eachindex(x)])
+    end
+
     @test repr(FunctionTransform(sin)) == "Function Transform: $(sin)"
     f(a, x) = sin.(a .* x)
     test_ADs(x -> SEKernel() ∘ FunctionTransform(y -> f(x, y)), randn(rng, 3))

test/transform/selecttransform.jl

@@ -7,8 +7,9 @@
     x_vecs = [randn(rng, maximum(select)) for _ in 1:3]
     x_cols = ColVecs(randn(rng, maximum(select), 6))
     x_rows = RowVecs(randn(rng, 4, maximum(select)))
+    x_string = ColVecs([randstring(rng) for _ in 1:maximum(select), _ in 1:5])
 
-    Xs = [x_vecs, x_cols, x_rows]
+    Xs = [x_vecs, x_cols, x_rows, x_string]
 
     @testset "$(typeof(x))" for x in Xs
         x′ = map(t, x)
@@ -24,8 +25,9 @@
     a_vecs = map(x -> AxisArray(x; col=symbols), x_vecs)
     a_cols = ColVecs(AxisArray(x_cols.X; col=symbols, index=(1:6)))
     a_rows = RowVecs(AxisArray(x_rows.X; index=(1:4), col=symbols))
+    a_string = ColVecs(AxisArray(x_string.X; col=symbols, index=1:5))
 
-    As = [a_vecs, a_cols, a_rows]
+    As = [a_vecs, a_cols, a_rows, a_string]
 
     @testset "$(typeof(a))" for (a, x) in zip(As, Xs)
         a′ = map(ts, a)
@@ -122,8 +124,9 @@
             ("Vector{<:Vector}", [randn(6) for _ in 1:3]),
             ("ColVecs", ColVecs(randn(5, 10))),
             ("RowVecs", RowVecs(randn(11, 4))),
+            ("ColVecs{String}", ColVecs([randstring() for _ in 1:6, _ in 1:5])),
         ]
-            @test KernelFunctions._map(t, x) isa AbstractVector{Float64}
+            @test KernelFunctions._map(t, x) isa AbstractVector
         end
     end
 end